Interpretation and Use of Applied/Operational Machine Learning and Artificial Intelligence in Surgery

Area of Science:

• Surgical Technology
• Artificial Intelligence in Medicine
• Machine Learning Applications

Background:

• Artificial intelligence (AI) and machine learning (ML) offer numerous surgical applications, including image interpretation, risk prediction, and robotic assistance.
• The rapid advancement of AI/ML algorithms in surgery contrasts with slower progress in demonstrating clinical utility, validity, and equity.

Purpose of the Study:

• To review the current state of AI and ML applications in surgery.
• To identify barriers to the widespread clinical adoption of AI in surgical practice.
• To propose a path forward for developing equitable and effective AI surgical systems.

Main Methods:

• Literature review of AI and ML applications in surgery.
• Analysis of challenges hindering clinical integration.
• Discussion of potential solutions involving multidisciplinary collaboration.

Main Results:

• AI/ML applications in surgery are diverse and rapidly evolving.
• Significant gaps exist between algorithm development and proven clinical utility, validity, and equity.
• Outdated infrastructure and regulatory hurdles create data silos, impeding AI progress.

Conclusions:

• Widespread adoption of AI in surgery is limited by challenges in demonstrating clinical utility, validity, and equity.
• Overcoming barriers requires addressing outdated computing infrastructure and regulatory issues.
• Multidisciplinary teams are essential for creating relevant, equitable, and dynamic AI surgical systems.